You are on page 1of 41

Chapter 5 Introduction to Factorial

Designs

1
5.1 Basic Definitions and Principles
Study the effects of two or more factors.
Factorial designs
Crossed: factors are arranged in a factorial design
Main effect: the change in response produced by a
change in the level of the factor

2
Definition of a factor effect: The change in the mean response when
the factor is changed from low to high

40 52 20 30
A y A y A 21
2 2
30 52 20 40
B yB yB 11
2 2
52 20 30 40
AB 1 3
2 2
50 12 20 40
A y A y A 1
2 2
40 12 20 50
B yB yB 9
2 2
12 20 40 50
AB 29
2 2
4
Regression Model &
The Associated
Response Surface

y 0 1 x1 2 x2
12 x1 x2
The least squares fit is
y 35.5 10.5 x1 5.5 x2
0.5 x1 x2
35.5 10.5 x1 5.5 x2

5
The Effect of
Interaction on the
Response Surface
Suppose that we add an
interaction term to the
model:

y 35.5 10.5 x1 5.5 x2


8 x1 x2

Interaction is actually
a form of curvature

6
When an interaction is large, the corresponding
main effects have little practical meaning.
A significant interaction will often mask the
significance of main effects.

7
5.2 The Advantage of Factorials
One-factor-at-a-time desgin
Compute the main effects of factors
A: A+B- - A-B-
B: A-B- - A-B+
Total number of experiments: 6
Interaction effects
A+B-, A-B+ > A-B- => A+B+ is
better???
8
5.3 The Two-Factor Factorial Design
5.3.1 An Example
a levels for factor A, b levels for factor B and n
replicates
Design a battery: the plate materials (3 levels) v.s.
temperatures (3 levels), and n = 4: 32 factorial design
Two questions:
What effects do material type and temperature have
on the life of the battery?
Is there a choice of material that would give
uniformly long life regardless of temperature? 9
The data for the Battery Design:

10
Completely randomized design: a levels of factor
A, b levels of factor B, n replicates

11
Statistical (effects) model:
i 1, 2,..., a

yijk i j ( )ij ijk j 1, 2,..., b
k 1, 2,..., n

is an overall mean, i is the effect of the ith level
of the row factor A, j is the effect of the jth
column of column factor B and ( )ij is the
interaction between i and j .
Testing hypotheses:
H 0 : 1 a 0 v.s. H 1 : at least one i 0

H 0 : 1 b 0 v.s. H 1 : at least one j 0

12
H 0 : ( ) ij 0 i, j v.s. H 1 : at least one ( ) ij 0
5.3.2 Statistical Analysis of the Fixed Effects
Model
b n
yi..
yi.. yijk yi..
j 1 k 1 bn
a n y. j .
y. j . yijk y. j .
i 1 k 1 an
n yij.
yij. yijk yij.
k 1 n
a b n
y...
y... yijk y...
i 1 j 1 k 1 abn

13
a b n a b

ijk ...
( y y )
i 1 j 1 k 1
2
bn i.. ...
( y y
i 1
) an . j. ...
( 2
y y ) 2

j 1
a b a b n
n ( yij . yi.. y. j . y... ) 2 ( yijk yij . ) 2
i 1 j 1 i 1 j 1 k 1

SST SS A SS B SS AB SS E
df breakdown:
abn 1 a 1 b 1 (a 1)(b 1) ab(n 1)

14
Mean squares
a
bn i2
E ( MS A ) E ( SS A /( a 1)) 2 i 1

a 1
b
an j2
j 1
E ( MS B ) E ( SS B /(b 1)) 2
b 1
a b
n ( ) ij2
SS AB i 1 j 1
E ( MS AB ) E ( ) 2
(a 1)(b 1) (a 1)(b 1)
SS E
E ( MS E ) E ( ) 2
ab(n 1)

15
The ANOVA table:

16
17
Example 5.1
Response: Life
ANOVA for Selected Factorial Model
Analysis of variance table [Partial sum of squares]
Sum of Mean F
Source Squares DF Square Value Prob > F
Model 59416.22 8 7427.03 11.00 < 0.0001
A 10683.72 2 5341.86 7.91 0.0020
B 39118.72 2 19559.36 28.97 < 0.0001
AB 9613.78 4 2403.44 3.56 0.0186
Pure E 18230.75 27 675.21
C Total 77646.97 35

Std. Dev. 25.98 R-Squared 0.7652


Mean 105.53 Adj R-Squared 0.6956
C.V. 24.62 Pred R-Squared 0.5826
PRESS 32410.22 Adeq Precision 8.178

18
DESIGN-EXPERT Plot Interaction Graph
Life A: Material
188

X = B: Temperature
Y = A: Material

A1 A1 146
A2 A2
A3 A3

Life
104

2
62
2

20

15 70 125

B: Tem perature

19
Multiple Comparisons:
Use the methods in Chapter 3.
Since the interaction is significant, fix the factor
B at a specific level and apply Turkeys test to
the means of factor A at this level.
See Page 174
Compare all ab cells means to determine which
one differ significantly

20
5.3.3 Model Adequacy Checking
Residual analysis: eijk yijk y ijk yijk yij
DESIGN-EXPERT Plot Normal plot of residuals DESIGN-EXPERT Plot
Life Life
Residuals vs. Predicted
45.25

99

95
18.75
90
Norm al % probability

80
70

Res iduals
50 -7.75

30
20

10 -34.25
5

-60.75

49.50 76.06 102.62 129.19 155.75


-60.75 -34.25 -7.75 18.75 45.25

Predicted
Res idual

21
DESIGN-EXPERT Plot Residuals vs. Run
Life
45.25

18.75

Res iduals -7.75

-34.25

-60.75

1 6 11 16 21 26 31 36

Run Num ber

22
DESIGN-EXPERT Plot DESIGN-EXPERT Plot Residuals vs. Temperature
Life
Residuals vs. Material Life
45.25 45.25

18.75 18.75

Res iduals
Res iduals

-7.75 -7.75

-34.25 -34.25

-60.75
-60.75

1 2 3
1 2 3

Tem perature
Material

23
5.3.4 Estimating the Model Parameters
The model is
y ijk i j ( ) ij ijk
The normal equations:
a b a b
: abn bn i an j n ( ) ij y
i 1 j 1 i 1 j 1
b b
i : bn bn i n j n ( ) ij y i
j 1 j 1
a a
j : an n i an j n ( ) ij y j
i 1 i 1

( ) ij : n n i n j n( ) ij y ij
Constraints: a b a b

i 0, j 0, ij ij 0
i 1 j 1 i 1 j 1
24
Estimations:
y
i yi y
j y j y
ij yij y i y j y

The fitted value:


y ijk i j ij yij

Choice of sample size: Use OC curves to choose


the proper sample size.
25
Consider a two-factor model without interaction:
Table 5.8
The fitted values: y ijk y i y j y

26
One observation per cell:
The error variance is not estimable because the
two-factor interaction and the error can not be
separated.
Assume no interaction. (Table 5.9)
Tukey (1949): assume ()ij = rij (Page 183)
Example 5.2

27
5.4 The General Factorial Design
More than two factors: a levels of factor A, b
levels of factor B, c levels of factor C, , and n
replicates.
Total abc n observations.
For a fixed effects model, test statistics for each
main effect and interaction may be constructed by
dividing the corresponding mean square for effect
or interaction by the mean square error.

28
Degree of freedom:
Main effect: # of levels 1
Interaction: the product of the # of degrees of
freedom associated with the individual
components of the interaction.
The three factor analysis of variance model:
yijkl i j k ( ) ij
( ) ik ( ) jk ( ) ijk ijkl
The ANOVA table (see Table 5.12)
Computing formulas for the sums of squares
(see Page 186)
Example 5.3
29
30
Example 5.3: Three factors: the percent
carbonation (A), the operating pressure (B); the
line speed (C)

31
32
5.5 Fitting Response Curves and
Surfaces
An equation relates the response (y) to the factor
(x).
Useful for interpolation.
Linear regression methods
Example 5.4
Study how temperatures affects the battery life
Hierarchy principle

33
Involve both quantitative and qualitative factors
This can be accounted for in the analysis to produce
regression models for the quantitative factors at each
level (or combination of levels) of the qualitative
factors
A = Material type
B = Linear effect of Temperature
B2 = Quadratic effect of
Temperature
AB = Material type TempLinear
AB2 = Material type - TempQuad
B3 = Cubic effect of
Temperature (Aliased)
34
35
36
37
5.6 Blocking in a Factorial Design
A nuisance factor: blocking
A single replicate of a complete factorial
experiment is run within each block.
Model:
yijk i j ( ) ij k ijk
No interaction between blocks and treatments
ANOVA table (Table 5.20)

38
39
Example 5.6:
Two factors: ground clutter and filter type
Nuisance factor: operator

40
Two randomization restrictions: Latin square
design
An example in Page 200.
Model:
yijkl i j k ( ) jk l ijkl
Tables 5.23 and 5.24

41

You might also like